Article Text

Download PDFPDF

‘Van Gogh’ syndrome: a term to approach with caution
  1. Brian Murray
  1. Older Adult Mental Health, Oxford Health NHS Foundation Trust, Oxford, Oxfordshire, UK
  1. Correspondence to Dr Brian Murray; brian.murray{at}

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

Mudgal et al provided us with an interesting and comprehensive case report.1 The subject of their report presents with typical features of schizophrenia, evidenced by his score on the BPRS (Brief Psychosis Rating Scale). The only atypical features are an age of onset which is slightly younger than average, but this could be explained by his use of cannabis (an estimate of usage would be helpful). Importantly, given the patient’s history of self-harm, Mudgal et al consider and rule out premorbid impulsive and borderline personality traits. Although the self-harm seems linked to command hallucinations, it would have been interesting to see more detail on the patient’s insight into this extreme behaviour. The degree of self-harm has led the authors to use the term ‘Van Gogh syndrome’.

‘Van Gogh syndrome’ is not in the ICD-10 (International Classification of Disease) nor DSM-V (Diagnostic and Statistical Manual). It is defined not in the medical literature but on Wikipedia, where it is considered a synonym for NSSI (Non-suicidal Self-Injury). NSSI requires five or more days of mild to moderate self-harm in a year. Self-harm as a result of psychosis is excluded. Mudgal et al, however, follow the psychiatric literature in reserving ‘Van Gogh syndrome’ for severe self-harm (usually mutilation) associated with psychosis: a review of the literature found five papers (including Mudgal et al’s), describing six patients with self-mutilation. All were diagnosed with psychosis: one with psychosis unspecified,2 three with schizophrenia1 3 4 and two with bipolar disorder.5 In common usage, therefore, Van Gogh syndrome does not equate to NSSI. The erratic nature of reporting means we cannot rule out selection bias: in other words, the common assumption that Van Gogh was psychotic (repeated by the Wikipedia article) seems to have created an expectation in the literature that the term ‘Van Gogh’ syndrome is reserved for psychotic patients who indulge in extreme self-harm.

There is a question, therefore, whether Van Gogh syndrome is a discrete condition. My medical dictionary describes a syndrome as ‘a group of signs or symptoms occurring together significantly often’. The importance of a causal explanation is debated, but many conditions with known causes are still called syndromes. Shown a child with learning difficulties, gout, choreoathetosis and self-mutilation, and the astute clinician will run tests for urate overproduction and the presence of HPRT1 (hypoxanthine phosphoribosyltransferase 1) gene in order to confirm Lesch-Nyhan syndrome. The concept of Lesch-Nyhan syndrome, therefore, has ‘tight’ definition (evidenced by a number of clear clinical signs/symptoms), plus predictive value and clinical utility.

If we follow the literature’s assumption that Van Gogh syndrome describes a specific syndrome linking psychosis to self-mutilation, does this carry the clarity and clinical utility of other syndromes? It is true that self-harm is more common in psychosis, but association between self-harm and a psychiatric diagnosis is far from unique. Other associations have been suggested for extreme self-harm in psychosis: a history of prior self-harm, radically altering one’s appearance, derogatory auditory hallucinations and object loss.6 Although useful to know, these are associations, rather than actual symptoms. Van Gogh syndrome currently remains a descriptive term rather than a clear nosological entity with genuine clinical meaning.

I would further argue that it is dangerous to base a ‘syndrome’ not on symptoms but on the vagaries of human behaviour. To be a ‘syndrome’ proper, Van Gogh syndrome needs diagnostic markers that can distinguish it from the wide range of bizarre behaviours exhibited in psychosis.

In fact, it is not even clear that Van Gogh was psychotic when on 23 December 1888, he quarrelled with his housemate Gauguin, cutoff a part of his left ear and presented it to ‘Rachel’, a maid in a brothel who, allegedly, they had rowed over (she was not, as is often supposed, a sex worker herself). Van Gogh wrote lucid letters, with no evidence of thought disorder or delusions, to his brother Theo on the day of the incident and a few days afterwards. Contemporaneous accounts from his physician and Theo describe Van Gogh as distressed immediately after the incident, but there is no suggestion of psychosis. The idea only emerged later, in selected readings of one letter in which Van Gogh compared (but also contrasted) himself with other inmates at the asylum to which he was eventually admitted.7

Wikipedia lists the following as possible causes for Van Gogh’s behaviour: epilepsy, bipolar disorder, schizoaffective disorder, Meniere’s disease, lead or absinthe poisoning, porphyria, anxiety, syphilis, emotionally unstable personality disorder and sunstroke. There is little to prove matters either way and suggestions such as syphilis, Meniere’s disease, lead poisoning or porphyria remain speculative. Van Gogh was said to have ‘epilepsy’, but this was a much broader term then than it is now. The absinthe of the day used much lower concentrations of thujone, and was therefore a lot safer, than is popularly thought.8 ‘Sunstroke’ is particularly implausible as the incident happened in winter and it had been raining for 3 days!

If any theory can be held up, it is probably that of emotionally unstable personality disorder: Van Gogh was perhaps the ultimate tortured artist. He cut an intense figure, lurching from relationship to relationship and job to job. He had a longstanding tendency to push and punish himself. His actions that fateful night may have had some purpose and meaning in the context of his argument with Gauguin, who he was terrified would leave him and his grand plans for an artistic community in Arles. The act of giving Rachel the ear may indicate Van Gogh blamed her in some way.9

Confusion over Van Gogh’s biography aside, if Van Gogh syndrome is indeed synonymous with NSSI, then authors would be advised to keep to the latter term and the clear criteria provided by DSM-V. If the literature continues to suggest that there is a specific syndrome of severe self-harm with psychosis, then this needs more detailed investigation to see if the syndrome has meaningful associations which can distinguish it from the spectrum of self-harm associated with mental illness. Van Gogh syndrome remains an evocative term, but its use should be avoided in favour of existing terminology approved by our current classification systems.

Finally, Wikipedia acknowledges that ‘Van Gogh’ syndrome can also be used to describe digoxin toxicity (again based on speculative biography) which further highlights the perils of eponymous syndromes!


Brian Murray obtained an MA degree in the Philosophy and Ethics of Mental Health. He gained his medical degree from Oxford University in 1992, working in general medical jobs before training in psychiatry in 1997. He has been a member of the Royal College of Psychiatrists since 2000 and a consultant psychiatrist since 2005. He is currently a Consultant Older Adult Psychiatrist at Oxford Health NHS Trust in the UK, and working as an inpatient consultant and has contributed to the Oxford Textbook of Inpatient Psychiatry. His main research interests include dementia and mental capacity.

Embedded Image


  • Funding The authors have not declared a specific grant for this research from any funding agency in the public, commercial or not-for-profit sectors.

  • Competing interests None declared.

  • Patient consent for publication Not required.

  • Provenance and peer review Commissioned; externally peer reviewed.